-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
463 lines (394 loc) · 26.1 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<link rel="stylesheet" href="styles.css" type="text/css">
<title>Yatong Bai</title>
</head>
<body>
<table summary="Table for page layout." class="tlayout">
<!-- Navigation bar on the left -->
<td id="layout-menu">
<div style="margin-bottom:4px"></div>
<div class="menu-item"><a href="index.html" class="current">Home</a></div>
<div class="menu-item"><a href="publications.html">Publications</a></div>
<div class="menu-item"><a href="teaching.html">Teaching</a></div>
<div class="menu-item"><a href="coursework.html">Coursework</a></div>
<div><hr></div>
<div class="menu-item"><a
href="https://scholar.google.com/citations?hl=en&user=A9h086cAAAAJ"
target=“blank”>Google Scholar</a></div>
<div class="menu-item"><a
href="https://www.semanticscholar.org/author/Yatong-Bai/2115297196"
target=“blank”>Semantic Scholar</a></div>
<div class="menu-item"><a href="https://www.linkedin.com/in/yatong-bai"
target=“blank”>LinkedIn</a></div>
<div class="menu-item"><a href="https://github.com/Bai-YT"
target=“blank”>GitHub</a></div>
<div class="menu-item"><a href="files/index/Resume Yatong Bai Two Pages.pdf"
target=“blank”>CV</a></div>
</td>
<!-- Main contents -->
<td id="layout-content">
<div id="toptitle">
<h1>Yatong Bai</h1>
</div>
<!-- Avatar photo and quick links -->
<table class="imgtable">
<td>
<img src="files/index/Photo.jpg" alt="Yatong Bai" width="260px" />
</td>
<td>
<section class="panel">
<p class="intro_large"><b>Ph.D. Candidate</b></p>
<div style="height:4px;"></div>
<p class="intro_large">University of California, Berkeley</p>
<p class="intro_large">
Advisor: <a href="https://people.eecs.berkeley.edu/~sojoudi/"
target=“blank”>Somayeh Sojoudi</a>
</p>
<p class="intro">
Email: <a href="mailto:yatong\_bai@berkeley.edu"
target=“blank”>yatong_bai (at) berkeley.edu</a>
</p>
<p class="intro">
<a href="https://bai-yt.github.io/files/index/Research Talk Open.pdf"
target=“blank”>Research Portfolio Slides</a>
</p>
</section>
</td>
</table>
<!-- Open to work -->
<!-- <section class="panel_highlight">
<i><b>I am looking for full-time research/engineering opportunities that start in 2025.</b></i>
</section> -->
<!-- About -->
<section class="panel2">
<h2>My Journey</h2>
<p>
I am a fifth-year Ph.D. candidate at the <a href="https://me.berkeley.edu"
target=“blank”>mechanical engineering</a> department of
<a href="https://www.berkeley.edu" target=“blank”>UC Berkeley</a>, advised by Professor
<a href="https://people.eecs.berkeley.edu/~sojoudi/" target=“blank”>Somayeh Sojoudi</a>.
</p>
<p>
Prior to joining Berkeley, I obtained Bachelor's degrees in computer engineering
and mechanical engineering from
<a href="https://www.gatech.edu" target=“blank”>Georgia Institute of Technology</a>,
where I researched with Professors
<a href="https://www.me.gatech.edu/faculty/meaud" target=“blank”>Julien Meaud</a>
and <a href=https://www.ece.gatech.edu/faculty-staff-directory/tom-conte
target=“blank”>Thomas Conte</a>.
</p>
<p>
I became increasingly interested in machine learning in my senior year, and decided to pursue
a Ph.D. in the field, focusing on efficient and reliable deep learning and generative AI.
</p>
<p>
I have interned at
<a href="https://www.adobe.com" target=“blank”>Adobe Research</a>,
<a href="https://www.microsoft.com" target=“blank”>Microsoft</a>,
<a href="https://scale.com" target=“blank”>Scale AI</a>,
<a href="https://www.hondajet.com" target=“blank”>Honda Aircraft Company</a>, and
<a href="https://www.tesla.com" target=“blank”>Tesla</a>.
</p>
</section>
<style>
p.intro {margin-bottom: 13px;} /* Adjust spacing below paragraphs */
li > p {margin-bottom: 8px;} /* Reduce space below section titles */
li ul {padding-left: 28px; margin-bottom: 5px;} /* Smaller indentation for nested lists */
li ul li {margin-bottom: 1px;} /* Space between sub-items */
</style>
<!-- Research insterests -->
<section class="panel2">
<h2>Research Interests</h2>
<p>
<b>My research aims to make deep learning and generative AI more efficient, aligned, robust, and reliable.</b>
My interests span across generative models (particulaly audio/music), robust deep learning,
(convex) optimization, reinforcement learning, and controls.
</p>
<p class="intro">
Specifically, my research areas so far include:
<ul>
<li><p> Diffusion Models (Audio/Music Generation). </p><ul>
<li> Accelerating diffusion-based text-to-audio generation with consistency distillation
(<a href="https://arxiv.org/abs/2309.10740" target=“blank”>ConsistencyTTA</a>). </li>
<li> Using reinforcement learning to align text-to-music generation to human preferences (ongoing). </li>
</ul></li>
<li><p> Robust Deep Learning. </p><ul>
<li> Analyzing the vulnerabilities of large language models (LLMs) when used in conversational search engines
(<a href="https://arxiv.org/abs/2406.03589" target=“blank”>RAGDOLL</a>). </li>
<li> Mixing classifiers to address the "accuracy-robustness trade-off" for image classifiers
(<a href="https://arxiv.org/abs/2402.02263" target=“blank”>MixedNUTS</a>,
<a href="https://arxiv.org/abs/2301.12554" target=“blank”>Adaptive Smoothing</a>). </li>
</ul></li>
<li><p> Convex Neural Network Optimization. </p><ul>
<li> Efficient algorithms for training two-layer ReLU neural networks with global optimality guarantees
(<a href="https://arxiv.org/abs/2201.01965" target=“blank”>link</a>). </li>
<li> Convex formulations for adversarial training, which build two-layer networks with adversarial robustness
(<a href="https://arxiv.org/abs/2105.12237" target=“blank”>link</a>). </li>
</ul></li>
</ul>
</p>
</section>
<!-- News -->
<section class="section">
<h2>News</h2>
<ul>
<li><p>
<b>January 2025:</b> I served as a reviewer for the
<a href="https://www.siam.org/publications/journals/siam-journal-on-mathematics-of-data-science-simods"
target=“blank”>SIMODS</a> journal.
</p></li>
<li><p>
<b>December 2024:</b> I lead the compilation of a knowledge summary about convex optimization based on
Berkeley's EECS 127/227A lecture notes by Somayeh Sojoudi. It's available to everyone
<a href="https://bai-yt.github.io/files/teaching/eecs127_notes/final_review.pdf" target=“blank”>here</a>.
</p></li>
<li><p>
<b>December 2024:</b> I served as a reviewer for the
<a href="https://jmlr.org/tmlr" target=“blank”>TMLR</a> journal.
</p></li>
<li><p>
<b>October 2024:</b> I served as a reviewer for the
<a href="https://iclr.cc/Conferences/2025" target=“blank”>ICLR 2025</a> conference.
</p></li>
<li><p>
<b>September 2024:</b> Our paper
<a href="https://arxiv.org/abs/2406.03589" target=“blank”>
Ranking Manipulation for Conversational Search Engines</a>
has been accepted to Conference on Empirical Methods in Natural Language Processing
<a href="https://2024.emnlp.org" target=“blank”>(EMNLP 2024)</a>
with an <b>Oral</b> presentation format (top 10% accepted papers).
During the November 12-16 conference in Miami, in addition to giving an oral presentation, we will
present <a href="files/publications/RAGDOLL_Poster.pdf" target=“blank”> this poster</a>.
</p></li>
<li><p>
<b>August 2024:</b> Our paper
<a href="https://arxiv.org/abs/2402.02263" target=“blank”>
“MixedNUTS: Training-Free Accuracy-Robustness Balance via Nonlinearly Mixed Classifiers”</a>
has been accepted to <a href="https://jmlr.org/tmlr"
target=“blank”>Transactions on Machine Learning Research (TMLR)</a>.
</p></li>
<li><p>
<b>July 2024:</b> I served as a reviewer for the
<a href="https://neurips.cc/Conferences/2024" target=“blank”>NeurIPS 2024</a> conference.
</p></li>
<li><p>
<b>June 2024:</b> New paper
<a href="https://arxiv.org/abs/2406.03589" target=“blank”>
“Ranking Manipulation for Conversational Search Engines”</a>
by Samuel Pfrommer, <b>Yatong Bai</b>, Tanmay Gautam, and Somayeh Sojoudi.
Project code <a href="https://github.com/spfrommer/ranking_manipulation"
target=“blank”>on GitHub</a>.
This work proposes the "RAGDOLL" dataset, which is available on
<a href="https://huggingface.co/datasets/Bai-YT/RAGDOLL"
target=“blank”>Huggingface Datasets</a>.
</p></li>
<li><p>
<b>June 2024:</b> Our paper
<a href="https://arxiv.org/abs/2309.10740" target=“blank”>
“ConsistencyTTA: Accelerating Diffusion-Based Text-to-Audio Generation with Consistency Distillation”</a>
has been accepted to <a href="https://interspeech2024.org"
target=“blank”>the INTERSPEECH conference of 2024</a>.
<b>Live demo available at <a href="https://huggingface.co/spaces/Bai-YT/ConsistencyTTA"
target=“blank”>Huggingface</a>. Feel free to try it out!</b>
</p></li>
<li><p>
<b>May 2024:</b> I have received the Henry Lurie Family Fellowship.
</p></li>
<li><p>
<b>March 2024:</b> Our paper
<a href="https://arxiv.org/abs/2301.12554" target=“blank”>
Improving the Accuracy-Robustness Trade-Off of Classifiers via Adaptive Smoothing</a> has been accepted
to <a href="https://www.siam.org/publications/journals/siam-journal-on-mathematics-of-data-science-simods"
target=“blank”>the SIAM Journal on Mathematics of Data Science (SIMODS)</a>.
</p></li>
<li><p>
<b>March 2024:</b> I served as a reviewer for the
<a href="https://icml.cc/Conferences/2024" target=“blank”>ICML 2024</a> conference.
</p></li>
<li><p>
<b>March 2024:</b> Our paper
<a href="https://arxiv.org/abs/2311.15165" target=“blank”>
Mixing Classifiers to Alleviate the Accuracy-Robustness Trade-Off</a>
has been accepted to <a href="https://l4dc.web.ox.ac.uk"
target=“blank”>the 6th Annual Learning for Dynamics & Control Conference</a>.
</p></li>
<li><p>
<b>February 2024:</b> We have open-sourced the training and evaluation code of the paper
“ConsistencyTTA: Accelerating Diffusion-Based Text-to-Audio Generation with Consistency Distillation”
at <a href="https://github.com/Bai-YT/ConsistencyTTA"
target=“blank”>github.com/Bai-YT/ConsistencyTTA</a>.
The model checkpoints are shared on <a href="https://huggingface.co/Bai-YT/ConsistencyTTA"
target=“blank”>Huggingface</a>.
</p></li>
<li><p>
<b>February 2024:</b> New paper
<a href="https://arxiv.org/abs/2402.02263" target=“blank”>
“MixedNUTS: Training-Free Accuracy-Robustness Balance via Nonlinearly Mixed Classifiers”</a>
by <b>Yatong Bai</b>, Mo Zhou, Vishal M. Patel, and Somayeh Sojoudi.
Project code <a href="https://github.com/Bai-YT/MixedNUTS"
target=“blank”>github.com/Bai-YT/MixedNUTS</a>.
</p></li>
<li><p>
<b>February 2024:</b> I prepared summarization notes about Nonlinear Optimization for Professor Javad Lavaei's
<a href="https://lavaei.ieor.berkeley.edu/Course_IEOR160_Fall_2023.html" target=“blank”>IEOR 160</a>
course and made it publicly available <a href="teaching.html" target=“blank”>on this website</a>.
</p></li>
<li><p>
<b>December 2023:</b> I will join <a href="https://research.adobe.com/"
target=“blank”>Adobe Research</a>
as a Research Intern in May 2024. I will be working with
<a href="https://research.adobe.com/person/nicholas-j-bryan/"
target=“blank”>Nicholas J. Bryan</a>.
</p></li>
<li><p>
<b>October 2023:</b> I served as a reviewer for the
<a href="https://iclr.cc/Conferences/2024" target=“blank”>ICLR 2024</a> conference.
</p></li>
<li><p>
<b>September 2023:</b> New paper
<a href="https://arxiv.org/abs/2309.10740" target=“blank”>
“ConsistencyTTA: Accelerating Diffusion-Based Text-to-Audio Generation with Consistency Distillation”</a>
by <b>Yatong Bai</b>, Trung Dang, Dung Tran, Kazuhito Koishida, and Somayeh Sojoudi.
This work accelerates text-to-audio generation by up to 400x.
Project website <a href="https://consistency-tta.github.io"
target=“blank”>consistency-tta.github.io</a>. Code is open-sourced.
</p></li>
<li><p>
<b>August 2023:</b> I served as a Graduate Student Instructor (GSI) for
<a href="https://lavaei.ieor.berkeley.edu/Course_IEOR160_Fall_2023.html"
target=“blank”>IEOR 160: Nonlinear and Discrete Optimization</a>
in the Fall 2023 semester.
</p></li>
<li><p>
<b>July 2023:</b> Our paper
<a href="https://arxiv.org/abs/2307.15980" target=“blank”>
Initial State Interventions for Deconfounded Imitation Learning</a>
has been accepted to <a href="https://cdc2023.ieeecss.org" target=“blank”>
the IEEE Conference on Decision and Control (CDC)</a>.
</p></li>
<li><p>
<b>June 2023:</b> I served as a reviewer for the
<a href="https://nips.cc/" target=“blank”>NeurIPS 2023</a> conference.
</p></li>
<li><p>
<b>April 2023:</b> I served as a reviewer for the <a href="https://cdc2023.ieeecss.org"
target=“blank”>2023 IEEE Conference on Decision and Control (CDC)</a>.
</p></li>
<li><p>
<b>February 2023:</b> I served as a reviewer for the
<a href="https://icml.cc/Conferences/2023" target=“blank”>ICML 2023</a> and
<a href="https://ieeeccta.org/" target=“blank”>IEEE CCTA 2023</a> conferences.
</p></li>
<li><p>
<b>January 2023:</b> New paper
<a href="https://arxiv.org/abs/2301.12554" target=“blank”>
“Improving the Accuracy-Robustness Trade-Off of Classifiers via Local Adaptive Smoothing”</a>
by <b>Yatong Bai</b>, Brendon G. Anderson, Aerin Kim, and Somayeh Sojoudi.
</p></li>
<li><p>
<b>January 2023:</b> New paper
<a href="https://arxiv.org/abs/2401.04575" target=“blank”>
“Let's Go Shopping (LGS) — Web-scale Image-text Dataset for Visual Concept Understanding”</a>
by <b>Yatong Bai</b>, Utsav Garg, Erhan Bas, Isidora Filipovic, Apaar Shanker,
Haoming Zhang, Samyak Parajuli, Amelia N. Chu, Eugenia D Fomitcheva,
Elliot Branson, Aerin Kim, Somayeh Sojoudi, and Kyunghyun Cho.
</p></li>
<li><p>
<b>December 2022:</b> I will join <a href="https://www.microsoft.com/applied-sciences"
target=“blank”>Microsoft Applied Science</a>
as a Research Intern in May 2023. I will be working with
<a href="https://www.microsoft.com/applied-sciences/people/kazuhito-koishida"
target=“blank”>Kazuhito Koishida</a>.
</p></li>
<li><p>
<b>December 2022:</b> I presented our poster on convex neural
network training at the year-end TBSI Symposium.
</p></li>
<li><p>
<b>October 2022:</b> Our paper
<a href="https://arxiv.org/abs/2201.01965" target=“blank”>
“Efficient Global Optimization of Two-layer ReLU Networks:
Adversarial Training and Quadratic-time Algorithms”</a> has been accepted to
<a href="https://www.siam.org/publications/journals/siam-journal-on-mathematics-of-data-science-simods"
target=“blank”> the SIAM Journal on Mathematics of Data Science (SIMODS)</a>.
</p></li>
<li><p>
<b>August 2022:</b> I am serving as a Graduate Student Instructor (GSI) for
<a href="https://lavaei.ieor.berkeley.edu/Course_IEOR160_Fall_2022.html"
target=“blank”>IEOR 160: Nonlinear and Discrete Optimization</a>
of the Fall 2022 semester.
</p></li>
<li><p>
<b>May 2022:</b> I have joined <a href="https://scale.com" target=“blank”>
Scale AI</a> as a Machine Learning Research Intern.
I will be working on a bi-modal dataset project with
<a href="https://github.com/aerinkim" target=“blank”>Aerin Kim</a>.
</p></li>
<li><p>
<b>May 2022:</b> I have passed the Qualifying Examination and advanced to candidacy as a Ph.D. candidate.
</p></li>
<li><p>
<b>April 2022:</b> I served as a reviewer for the <a href="https://cdc2022.ieeecss.org"
target=“blank”>2022 IEEE Conference on Decision and Control (CDC)</a>.
</p></li>
<li><p>
<b>January 2022:</b> Our paper
<a href="https://people.eecs.berkeley.edu/~sojoudi/ACC_YA_22.pdf" target=“blank”>
“Practical Convex Formulation of Robust Two-layer Neural Network Training”</a>
has been accepted to the American Control Conference of 2022.
</p></li>
<li><p>
<b>January 2022:</b> I am serving as a Graduate Student Instructor (GSI) for
<a href="https://lavaei.ieor.berkeley.edu/Course_IEOR160_Spring_2022.html" target=“blank”>
IEOR 160: Nonlinear and Discrete Optimization</a>
of the Spring 2022 semester.
</p></li>
<li><p>
<b>December 2021:</b> New paper
<a href="https://arxiv.org/abs/2201.01965" target=“blank”>
“Efficient Global Optimization of Two-layer ReLU Networks:
Adversarial Training and Quadratic-time Algorithms”</a>
by <b>Yatong bai</b>, Tanmay Gautam, and Somayeh Sojoudi.
</p></li>
<li><p>
<b>October 2021:</b> I gave a talk on convex neural network adversarial training at INFORMS 2021.
</p></li>
<li><p>
<b>October 2021:</b> The paper “Efficient Global Optimization of Two-layer ReLU Networks:
Adversarial Training and Quadratic-time Algorithms” received the 2021 <a href="https://www.informs.org/"
target=“blank”>INFORMS</a> Data Mining Best Student Paper Award Runner-Up (2nd out of 48 papers).
</p></li>
<li><p>
<b>August 2021:</b> I gave a talk on convex neural network adversarial training at MOPTA 2021.
</p></li>
<li><p>
<b>August 2021:</b> I have passed the Preliminary Examination.
</p></li>
<li><p>
<b>April 2021:</b> I received the ECE Senior Scholar Award from the
<a href=https://www.ece.gatech.edu/news/646699/ece-faculty-staff-and-students-receive-honors-roger-p-webb-awards-program
target=“blank”>Roger P. Webb Awards Program</a> at Georgia Tech.
</p></li>
<li><p>
<b>December 2020:</b> New paper on convex neural network adversarial training:
<a href="https://people.eecs.berkeley.edu/~sojoudi/ACC_YA_22.pdf" target=“blank”>
“Practical Convex Formulation of Robust Two-layer Neural Network Training”</a>
by <b>Yatong Bai</b>, Tanmay Gautam, Yu Gai, and Somayeh Sojoudi.
</p></li>
<li><p>
<b>August 2020:</b> I have joined UC Berkeley as a Ph.D. student.
</p></li>
<li><p>
<b>August 2020:</b> I have graduated from Georgia Institute of Technology with B. S. degrees in
Computer Engineering and Mechanical Engineering with a cumulative GPA of 4.00 out of 4.00.
</p></li>
</ul>
</section>
</td>
</table>
</body>
</html>